Processing Method Research Based On The Characteristic Of Spectral Image

Imaging spectrometer, as one device of the earth-observing systems, is an important embodiment of the development direction of the remote sensing technique, and a remarkable achievement in it's development. It integrates spectrum and imaging, having the ability of obtaining high spatial and spectral resolution hyperspectral images, and a wide range of applications. Therefore it will become a major optical remote-sensing device in the next twenty years.Firstly, this paper briefly introduces the development of imaging spectrometer, and the recent development of the imaging spectral technique.Secondly, it describes the fundamental theory of image's correlation and entropy, and analyses the characteristics of the imaging spectrometer images in detail, points out spectral image's substantial characteristics of the highly spatial and spectral correlation, which are different from other images'. In the respect' of the . entropy analysis, it gives the lower limit of spectral image's entropy from the single and the all band spectral images grouping, using the Gauss and Laplacian noise probability density distributing model. The lower limit of entropy is the lowest bit rate of the image loss-less compression.Thirdly, it studies the image's strip correction with the correlation characteristics of the imaging spectrometer images, and advances the surface "fitting correction method for the imaging spectrometer images. The correction method greatly reduces the correcting process influence upon spectral characteristic information, and improves images correcting quality.Fourthly, it discusses the spectral image loss-less compression by applying thecorrelation and the entropy of the spectral image sequence. It gives and analyses three expressions of imaging spectrometer images, and tells the change in correlation coefficient and entropy for different expressions. At last, it puts forward D2PCM de-correlation method to reduce the spatial and spectral redundancy. This method improves the compression ratio and the compression efficiency, being easy and fast in operation.